6.p = unite weight mean square error for height coor- 
dinates : 
6, 6y,0z= mean Square errors of true error in check point 
coordinates 
6p = mean square error of true error of planimetric co- 
ordinates at, check points» then 
Gp = +{(6x + Gy)2)h 
T is the relative accuraoy of heizhts at check points 
All values as above are ones in photo scale units:wm 
1. Relationship between effect on a posteriori compensation 
and Signal-to-noise ratio 
In order to define the relationship between effect on a 
posteriori compensation and signal-to-noise ratio, the first 
thing for us to do is to define signal-to-noise ratio. As 
this kind of systematic error is added to the simulated data, 
this svstematic errors will cause unequal distortions to dif- 
ferent parameters in computational results including Goes Gop + 
Ge » G2 etce Therefore, it is necessary to define their signal- 
to-noise ratio individually. It can be confirmed that, after the 
addition of systematic distortion, mean square errors thus ob- 
tained without using a posteriori compensation are results 
under the common influence of systematic and random errors; 
that is; 
= 2 2 
Gr ==Gg +0a 
Computational results without adding systematic errors and 
using a posteriori compensation are only results under the 
influence of random errors, So signal-to-noise ratio can be 
obtained bv 
2 1% 
Gs (Gr- a ) 
C= — = 
Ga Ga 
  
According to the above formula, the signal-to-noise ratio 
thus obtained is shown in Table 2. In Table 2, accuracy in- 
crease of rate of multiple after a posteriori compensation 
is listed statisticallv. 
In order to understand more clearly the benefit of a 
- 495 -